US20130155912A1 - Multiple-Input Multiple-Output Wireless Communications with Full Duplex Radios - Google Patents

Abstract

An apparatus used in a full duplex multiple-input multiple-output (MIMO) wireless communications system is disclosed. In an aspect, the apparatus includes a plurality of antennas, each of which is connected to a combination of at least a TX (transmit) RF (radio frequency) chain and an RX (receive) RF chain. The plurality of antennas are split into one or more transmit antennas and one or more receive antennas, and the apparatus communicates with another apparatus. Other apparatuses and some methods for wireless communications also are disclosed.

Description

• This application claims the benefit of U.S. Provisional Application No. 61/531,057, entitled, “Methods and apparatus for single cell and single link full duplex communications,” filed Sep. 5, 2011, the contents of which are incorporated herein by reference.
BACKGROUND OF THE INVENTION
• The present invention relates to multiple-input multiple-output (MIMO) wireless communications and more particularly to MIMO wireless communications with full duplex radios.
• Realization of the full duplex communication systems requires overcoming multiple implementation challenges. In particular it is very important to realize a system that can use full duplex communication without sacrificing the benefits of multiple antenna technologies. A practical system may address (1) how many antennas should be assigned for reception and transmission, respectively, (2) in OFDM (Orthogonal Frequency Division Multiplexing) systems how the assignment of the uplink and downlink should be performed, (3) in a single cell MIMO systems how the base station should schedule the uplink and downlink users and what should be the power split, and (4) finally in asynchronous single cell systems how should the MAC layer be designed to exploit the full potential of the full duplex access point as well as the full duplex clients.
• Some prior works have considered the design of full duplex communication systems:
• [1] M. Jain, J. Choi, T. Kim, D. Bharadia, S. Seth, K. Srinivasan, P. Levis, S. Katti, and P. Sinha, “Practical, real-time, full duplex wireless,” 2011.
• [2] A. Sahai, G. Patel, and A. Sabharwal, “Pushing the limits of full-duplex: Design and real-time implementation,” Arxiv preprint arXiv:1107.0607, 2011.
• [3] B. Radunovic, D. Gunawardena, A. Proutiere, N. Singh, V. Balan, and P. Key, “Efficiency and fairness in distributed wireless networks through self-interference cancellation and scheduling,” Tech. Rep. MSR-TR-2009-27, Microsoft Research, March 2009, http://research. microsoft. com/apps/pubs/default. aspx, Tech. Rep.
• [4] S. Rangarajan, X. Zhang, S. Barghi, M. A. Khojastepour, and K. Sundaresan, “The case for antenna cancellation for scalable full-duplex wireless communications,” Tech. Rep. 2011-TR074, NEC Laboratories America, Inc., Tech. Rep.
• [5] W. Pradeep Chathuranga, C. Marian, L. Matti, and E. Anthony, “On the effect of self-interference cancelation in multihop wireless networks,” EURASIP Journal on Wireless Communications and Networking, vol. 2010, 2010.
• In [2] the authors estimate the channel and reconstruct the self-interference from digital samples. By using an extra transmit antennas the authors in [1] create a null at a single receive antenna. A digital noise cancellation algorithm known as active noise cancelation is used in [1]. The implementation of the active noise cancellation is performed through the use of QHx220 chip. In prior work [4], the authors have proposed the use of two copies of the self-interference signal in which case we need an extra Receive antenna for each original receive antennas.
• In this specification, we address schemes and methods that address all the four problems described above. In particular, we provide guidelines on how to split the antennas between the transmit and receive RF (radio frequency) chains. We also provide method of allocating different tones in OFDM systems to uplink, downlink or full duplex (simultaneous uplink downlink) transmission. We also address the problem of user scheduling for full duplex communication in single cell. The full duplex scheduling is challenging due to the interferences that the uplink users will cause on the downlink users. The proposed scheduling algorithm depends on the number of transmit antennas and the number of active users and their channel gains. For a single cell systems when the scheduling is not possible and random access scheme is used, we also propose a method by which the base station or the access point can admit a new uplink communication while a downlink communication is in progress or it may initiate a new downlink transmission when an uplink transmission is already in session.
BRIEF SUMMARY OF THE INVENTION
• An objective of the present invention is to simultaneously send in the uplink and downlink by enabling full duplex communication.
• An aspect of the present invention includes an apparatus used in a full duplex MIMO wireless communications system. The apparatus includes a plurality of antennas, each of which is connected to a combination of at least a TX (transmit) RF (radio frequency) chain and an RX (receive) RF chain. The plurality of antennas are split into one or more transmit antennas and one or more receive antennas, and the apparatus communicates with another apparatus.
• Another aspect of the present invention includes a method implemented in a base station used in a full duplex MIMO wireless communications system. The method includes dividing a set of OFDM tones into a first group for uplink and a second group for downlink, scheduling a first user equipment in a first group and a second user equipment in the second group, receiving, from the second user equipment, information about interference on an OFDM tones in the first group, and scheduling one or more downlink users in the first group according to the information.
• Still another aspect of the present invention includes a method implemented in a base station used in a full duplex MIMO wireless communications system. The method includes dividing a set of OFDM tones into a first group for uplink and a second group for downlink, scheduling a first user equipment in a first group and a second user equipment in the second group, measuring downlink enhancement, and allocating one or more OFDM tones in the second group to uplink according to the downlink enhancement.
• Still another aspect of the present invention includes a method implemented in a base station used in a full duplex MIMO wireless communications system. The method includes scheduling one or more user equipment in an OFDM tone used for full duplex communications, according to a prior schedule and an achieved rate.
BRIEF DESCRIPTION OF THE DRAWINGS
• FIG. 1 depicts full duplex communications.
• FIG. 2 depicts strong self interference signals in comparison to weak received signals from an intended transmitter.
• FIG. 3 depicts antenna deployment scenarios with the same number of antennas and with the same number of transmit and receive RF chains.
• FIG. 4 depicts FD (full-duplex) gain with outage capacity as a measure.
• FIG. 5 depicts average gain of using FD radio with different antenna breakages, for N=8, M=4.
• FIG. 6 depicts FD gain (half-duplex clients, breaking the antennas at the AP (access point)) a) for different number of antennas at the AP, and b) for different number of clients.
• FIG. 7 depicts FD gain a) for different number of antennas at the AP and b) for different number of clients.
• FIG. 8 depicts FD gain (full-duplex clients) a) for different number of antennas at the AP and b) for different number of clients.
DETAILED DESCRIPTION
• By enabling full duplex communication we can simultaneously send in the uplink and downlink and it could mean doubling the use of spectrum (FIG. 1). In half-duplex systems we either receive or transmit in time TDD (time-division duplex) or in frequency FDD (frequency-division duplex), so it may be thought that we waste half of the resources.
• The main challenge of the full duplex communication is to cancel the self interference that is orders of magnitude stronger than the received signal from the intended transmitters. This interference is partly known due to the fact that the transmitter exactly knows its own transmitted signal, though the exact channel between the transmit and receive antennas at the base station is not known. It would be even worse if this channel is time varying or fading because we then need to estimate this channel more frequently. FIG. 2 shows the strong self interference in comparison to the weak received signal from a mobile station or user.
• There are two possible deployment of the full duplex communication with respect to a fixed number of transmit and receive RF chains 302, 304, 306, and 308. In practice the main complexity involved with the use of multiple antennas 310 and 312 is associated with the number of RF chain due to the fact that channel estimation, precoding, beamforming, multiple stream transmission, and demodulation all depends on the number of receive RF chains or transmit RF chains. Depending on if we use one antenna 330 for each pair of transmit RF chain 322 and receive RF chain 324 or if we use two antennas 350 and 354, which means one 354 for receive RF chain 346 and the other 350 for transmit RF chain 342, we can have one of the two possible deployment scenarios. See FIG. 3.
• Both systems may have marginal pros or cons in half duplex systems. We may use one antenna per RF chain for full duplex communication because it does not change the system complexity. However, the cost associated with using more physical antennas well worth the possible gain that can be achieved by this deployment scenario. In the sequel, we address how to allocate the antennas for either transmit or receive if this deployment scenario is used.
• 1) In single link system, if both links have one transmit antenna per RF chain there is no need for antenna splitting and the average gain of FD with respect to HD (half-duplex) transmission would be 2. However, for particular realization of the channel this gain could be different. If one antenna is used for a pair of TX (transmit) and RX (receive) RF chain, antennas can be split into one or more transmit antennas and one or more receive antennas. The antenna splitting between the transmit and receive antennas is very critical. The antenna splitting may be based on the channel condition. The system may perform the antenna splitting so that capacity gain is increased. If the system cannot dynamically change the antenna selection for full duplex communication very fast, then we may set the antenna splitting that will work in average. We show that the average capacity gain of full duplex with respect to half duplex is given by the following formula:
• $\varepsilon \left\{G_{\mathrm{FD}}\right\}\approx \{\begin{array}{cc}\frac{\left(M-M_1\right)+\left(N-N_1\right)}{0.5\left(M+N\right)}=2-2\frac{{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="US20130155912A1-20130620-D00007.png"\; state="\{\{state\}\}">N</figure-callout>}}_1+M_1}{N+M}& \rho <<1\\ \frac{\min \left({\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="US20130155912A1-20130620-D00007.png"\; state="\{\{state\}\}">N</figure-callout>}}_1,M-M_1\right)+\min \left(N-{\mathrm{<figure-callout\; id="1"\; label="N"\; filenames="US20130155912A1-20130620-D00007.png"\; state="\{\{state\}\}">N</figure-callout>}}_1,M_1\right)}{\min \left(N,M\right)}& \rho >>1\end{array}$
• where G_FD is the capacity gain, ε{ } is the expectation operator or statistical average, ρ represents the average SNR (signal to noise ratio), M and N are the numbers of antennas at node 1 and 2, respectively, and M₁ and N₁ are the numbers of transmit antennas at node 1 and node 2, respectively.
• From the above formula, it can be the best to set N₁ and M₁ to one regardless of the number of antennas N and M. This means that each node selects only one antenna for transmission and uses all the other antennas for reception. At high SNR, the split can be almost equal between the number of transmit and receive antennas at each node. In particular, if N=M, then M₁=N₁=M/2 maximizes the gain. If N<M then the solution is to set N₁=N/2. However, M₁ can take different values as long as M>M₁+N₁>N. M/2 and N/2 can be M/2±1/2 and N/2±1/2 when M and N are odd, respectively.
• 2) When OFDM system is used, one problem is how to find the splits between the OFDM tones into three groups where either one of uplink, downlink or simultaneous FD transmission is used. One strategy is to dynamically let the system finds the splits between these three groups. Since initiation of a full duplex communication in a single tone requires the knowledge of the interference in that tone on the downlink users, it is critical to have this information before scheduling any user in the downlink in that particular tone. In time varying environment it is important to periodically have this information before the scheduling. Therefore, the system begins the operation by dividing the set of OFDM tones into two groups: one for the uplink and the other one for the downlink. We assume that within a group of frames, e.g. a super frame, the channel does not change. While the system is in operation, the users in the downlink measure the interference on one or multiple OFDM tones that are in the uplink group. The downlink user then report a free channel signal in case that this interference is below a certain threshold. This threshold may be fixed or set periodically by BS (base station) or MS (mobile station). The initiation of the feedback of this information could be by the base station or by the MS. There may also be a fixed schedule when a downlink user may feedback this information.
• Based on this information, the BS then may select one or more downlink users to be scheduled in the same set of tones as some uplink users. This procedure allows for allocating some of the uplink tones to full duplex operation. This operation can only boost the downlink throughput. In order to allow for enhancement in the uplink, we can also have a secondary method in place. Over multiple super frames we measure the downlink enhancement and then we allocate some of the downlink tones to uplink. The enhancement may be measured in terms of the throughput, other quality of service or system utility measures. This reassignment of downlink tones to uplink would be dependent on the enhancement that we have received over a group of past superframes. By having both methods in place, it is then possible to exchange some of the enhancement received in the downlink for enhancement in the system utility for the uplink.
• 3) The scheduling of the users for the simultaneous uplink and downlink is based on an optimization of a system utility such as weighted sum rate where the weight itself changes in time based on the past operation of the systems and channel condition. The entire operation is discussed in the further system details.
• 4) Simultaneous uplink and downlink operation in the CSMA (carrier sense multiple access) single cell network also requires a modified MAC (media access control) layer.
• The foregoing is to be understood as being in every respect illustrative and exemplary, but not restrictive, and the scope of the invention disclosed herein is not to be determined from the Detailed Description, but rather from the claims as interpreted according to the full breadth permitted by the patent laws. It is to be understood that the embodiments shown and described herein are only illustrative of the principles of the present invention and that those skilled in the art may implement various modifications without departing from the scope and spirit of the invention. Those skilled in the art could implement various other feature combinations without departing from the scope and spirit of the invention.

Claims (18)

What is claimed is:

1. An apparatus used in a full duplex multiple-input multiple-output (MIMO) wireless communications system, the apparatus comprising:

a plurality of antennas, each of which is connected to a combination of at least a TX (transmit) RF (radio frequency) chain and an RX (receive) RF chain,

wherein the plurality of antennas are split into one or more transmit antennas and one or more receive antennas, and

wherein the apparatus communicates with another apparatus.

2. The apparatus as in claim 1,

wherein the antenna splitting is based on a channel condition.

3. The apparatus as in claim 1,

wherein the antenna splitting is performed so that capacity gain is increased.

4. The apparatus as in claim 3,

wherein the capacity gain is given by the following formula:

$\varepsilon \left\{G_{\mathrm{FD}}\right\}\approx \{\begin{array}{cc}\frac{\left(M-M_1\right)+\left(N-N_1\right)}{0.5\left(M+N\right)}=2-2\frac{N_1+M_1}{N+M}& \rho <<1\\ \frac{\min \left(N_1,M-M_1\right)+\min \left(N-N_1,M_1\right)}{\min \left(N,M\right)}& \rho >>1\end{array}$

where G_FD is the capacity gain, ε{ } is an expectation operator or statistical average, ρ represents an average SNR (signal to noise ratio), M and N are the numbers of antennas at the apparatus and said another apparatus, respectively, and M₁ and N₁ are the numbers of transmit antennas at the apparatus and said another apparatus, respectively.

5. The apparatus as in claim 1,

wherein, in case that an SNR (signal to noise ratio) is less than a predetermined value, N₁ and M₁ are set to one, where M₁ and N₁ are the numbers of transmit antennas at the apparatus and said another apparatus, respectively.

6. The apparatus as in claim 1,

wherein, in case that an SNR (signal to noise ratio) is more than a predetermined value, M₁ is set to M/2 (when M is even) or M/2±1/2 (when M is odd) and N₁ is set to N/2 (when N is even) or N/2±1/2 (when N is odd), where M and N are the numbers of antennas at the apparatus and said another apparatus, respectively, and M₁ and N₁ are the numbers of transmit antennas at the apparatus and said another apparatus, respectively.

7. The apparatus as in claim 1,

wherein, in case that an SNR (signal to noise ratio) is more than a predetermined value, M₁ is set to a value so that the following in equation is satisfied:

M>+N ₁>N,

where M and N are the numbers of antennas at the apparatus and said another apparatus, respectively, and M₁ and N₁ are the numbers of transmit antennas at the apparatus and said another apparatus, respectively.

8. A method implemented in a base station used in a full duplex multiple-input multiple-output (MIMO) wireless communications system, the method comprising:

dividing a set of OFDM (Orthogonal Frequency Division Multiplexing) tones into a first group for uplink and a second group for downlink;

scheduling a first user equipment in a first group and a second user equipment in the second group;

receiving, from the second user equipment, information about interference on an OFDM tones in the first group; and

scheduling one or more downlink users in the first group according to the information.

9. The method as in claim 8, wherein the information comprises a report on a free channel signal in case that the interference is below a threshold.

10. The method as in claim 9, wherein the threshold is fixed or set by the base station or the first user equipment.

11. The method as in claim 8, wherein the information feedback is initiated by the base station or by the first user equipment.

12. A method implemented in a base station used in a full duplex multiple-input multiple-output (MIMO) wireless communications system, the method comprising:

dividing a set of OFDM (Orthogonal Frequency Division Multiplexing) tones into a first group for uplink and a second group for downlink;

scheduling a first user equipment in a first group and a second user equipment in the second group;

measuring downlink enhancement; and

allocating one or more OFDM tones in the second group to uplink according to the downlink enhancement.

13. The method as in claim 12, wherein the downlink enhancement is measured in terms of at least one of throughput, quality of service, and a system utility measure.

14. The method as in claim 12, wherein the downlink enhancement is measured over one or more superframes, each of which comprises a group of frames.

15. A method implemented in a base station used in a full duplex multiple-input multiple-output (MIMO) wireless communications system, the method comprising:

scheduling one or more user equipment in an OFDM (Orthogonal Frequency Division Multiplexing) tone used for full duplex communications, according to a prior schedule and an achieved rate.

16. The method as in claim 15, wherein the achieved rate comprises a weighted sum rate.

17. The method as in claim 16, wherein a weight for the weighted sum rate changes in time according to at least one of operation of the full duplex MIMO wireless communications system and a channel condition.

18. The method as in claim 16, wherein the weighted sum rate can be expressed as the following expression:

$\max \sum _{i=1}^n\frac{1}{{\mathrm{Ra}}_i\left(t\right)}R_i\left(t\right),\mathrm{where}$ ${\mathrm{Ra}}_i\left(t+1\right)=\left(1-\frac{1}{T_c}\right){\mathrm{Ra}}_i\left(t\right)+\frac{1}{T_c}R_i\left(t\right),$

Ra_i is an average rate for link i, T_c is an averaging window size, and R_i(t) is a rate associated with link i in time-slot t.

Images